report of the iccat gbyp aerial survey for bluefin ... · available in 2017, always using the...
TRANSCRIPT
SCRS/2017/149 Collect. Vol. Sci. Pap. ICCAT, 74(6): 3172-3204 (2018)
3172
REPORT OF THE ICCAT GBYP AERIAL SURVEY FOR
BLUEFIN SPAWNING AGGREGATIONS IN 2017
A. Di Natale1, A. Cañadas2, J.A. Vázquez-Bonales2, S. Tensek1, A. Pagá García1
SUMMARY
The fifth ICCAT GBYP aerial survey was carried out in 2017, over 4 areas, overlapping with the
corresponding areas in previous surveys. It was carried out during the peak of the Bluefin tuna
spawning period (mostly in June), by 3 companies and 4 aircrafts. A new survey design was
available in 2017, always using the DISTANCE methodology, adopting an updated protocol. The
survey reports were provided in real-time and the survey ended on the 3rd July and therefore the
data analyses have been available for the first year in real time, according to what was set by the
ToRs of the contract. The elaboration provides the estimates for each area (number of schools,
number of tunas and quantities), with the usual details, comparing the results with those obtained
in previous years in the same areas. The results show a high interannual variability and a high
potential SSB. We detected in real-time an important shifting in the abundance between areas. It
is again evident that a reliable trend can be obtained only when considering various areas
together.
RÉSUMÉ
La cinquième prospection aérienne de l'ICCAT-GBYP a été réalisée en 2017, sur quatre zones,
couverture chevauchant avec les zones correspondantes des prospections antérieures. Celle-ci a
été réalisée au point culminant de la période de frai du thon rouge (surtout en juin), par trois
sociétés qui ont utilisé quatre aéronefs. Une nouvelle conception de la prospection a été
disponible en 2017, utilisant toujours la méthodologie DISTANCE, adoptant un protocole mis à
jour. Les rapports de prospection ont été fournis en temps réel et la prospection a pris fin le 3
juillet et les analyses des données ont par conséquent été disponibles pour la première année en
temps réel, conformément à ce qui avait été établi dans les termes de référence du contrat.
L'élaboration donne les estimations pour chaque zone (nombre de bancs, nombre de thons et
quantités), avec les détails habituels, en comparant les résultats avec ceux obtenus au cours des
années précédentes dans les mêmes zones. Les résultats montrent une forte variabilité
interannuelle et une SSB potentielle élevée. Nous avons détecté en temps réel un changement
important dans l’abondance entre les zones. Il est une fois de plus manifeste qu’une tendance
fiable peut être obtenue uniquement si l’on tient conjointement compte de diverses zones.
RESUMEN
La quinta prospección aérea del ICCAT-GBYP se llevó a cabo en 2017 sobre cuatro zonas,
solapándose con las zonas correspondientes a prospecciones anteriores. Fue llevada a cabo
durante el pico del periodo de desove del atún rojo (principalmente en junio) por tres empresas
que utilizaron cuatro aeronaves. En 2017 se dispuso de un nuevo diseño de la prospección, se
siguió utilizando la metodología DISTANCE y se adoptó un protocolo actualizado. Los informes
de la prospección se proporcionaron en tiempo real, y la prospección finalizó el 3 de julio y, por
lo tanto, es el primer año en el que los análisis están disponibles en tiempo real, de conformidad
con lo que se estableció en los términos de referencia del contrato. La elaboración proporciona
las estimaciones para cada área (número de bancos, número de túnidos y cantidades) con los
detalles habituales, comparando los resultados con los obtenidos en años anteriores en las
mismas áreas. Los resultados muestran una elevada variabilidad interanual y una SSB
potencialmente alta. Se detectó en tiempo real un cambio importante en la abundancia entre las
zonas. Una vez más resulta evidente que puede obtenerse una tendencia fiable solo si se
consideran varias zonas juntas.
1 ICCAT GBYP, Calle Corazón de María, 8, 6º, 28002 Madrid, Spain 2 Alnilam Investigación y Conservación S.L., Calle Pradillos 29, 28491 Navacerrada, Spain.
3173
KEYWORDS
Bluefin tuna, ICCAT, aerial survey, species distribution, Mediterranean, methodology
1. Foreword
Aerial survey are used for obtaining fishery independent data for some marine species or for more closely studying
their behaviour (Heldt, 1932; Grierson, 1949; Cram and Hapton, 1976; Rivas, 1978; Arena et al., 1979; Arena,
1980, 1981, 1982a, 1982b, 1982c, 1985, 1986a, 1986b, 1988a, 1988b, 1990; Marsh and Sinclair, 1989; Cowling
et al., 1996; Polacheck et al., 1996; Lutcavage et al., 1997; Hiby and Lovell, 1998; Cowling and O’Reilly, 1999;
Lutcavage and Newland, 1999; Buckland et al., 2001; Fromentin, 2001; Arena and Cefali, 2002; Hammond et al.,
2002; Thomas et al., 2002; Fromentin et al., 2003, 2013; Nicholson and Jennings, 2004; Newlands et al., 2007:
Bonhommeau et al., 2010; Everson et al., 2011; Farley and Bennet, 2011; Fortuna et al., 2011, 2014; Lauriano et
al., 2011, 2017; Palka, 2011; Panigada et al., 2011, 2017; Kessel et al., 2013; Basson and Farley, 2014; Bower et
al., 2014; Bauer et al., 2015a, 2015b; Rouyer 2016, in press), not in all cases appropriately using this technique or
not always providing the necessary details.
The ICCAT GBYP aerial survey for Bluefin tuna spawning aggregations is a method for obtaining fishery
independent indices of the Bluefin tuna spawning stock biomass over the years and, therefore, for possibly
obtaining trends, taking into account the implicit variability and the additional variance due to many factors; it is
implicit that estimates from the aerial survey are in the best case the minimum estimates, because they reflect the
quantities really encountered, which are always much less than the real quantity of fish, due to many natural factors
(mostly for the vertical distribution of the schools). From a management point of view, this represents a sort of
precautionary point of view. The initial decision for carrying on the survey on spawners and not on juveniles was
taken by the SCRS and confirmed by a SWOT analysis (Di Natale and Idrissi, 2013b). The previous surveys were
carried out in 2010 (Di Natale, 2011), 2011 (Di Natale and Idrissi, 2012, 2013a), 2013 (Di Natale et al., 2014a,
2014b) and 2015 (Di Natale and Tensek, 2016; Di Natale et al., 2016), depending on the availability of funds and
the choices of the GBYP Steering Committee, the SCRS and the Commission. All results and reports are available
on the ICCAT GBYP web pages http://www.iccat.int/GBYP/en/asurvey.htm .
The previous four ICCAT GBYP surveys were carried out with yearly changes, set by the GBYP Steering
Committee. The plan set by the SCRS and approved by the Commission at the beginning of GBYP was to survey
three areas for three years, but this plan was not sufficient for detecting any trend, as it was revealed later by a
power analysis requested by the Steering Committee (Cañadas and Vázquez, 2013), where a minimum of 6 years
was necessary. The total original budget set for three surveys in three areas was 1,200,000 Euros; the cost of
carrying out the previous four surveys in many more areas (four main “internal” areas and seven “external” areas)
was approximately 1,619,600 Euros (134.97% of the original budget, but with much more than twice the
activities); the aerial survey workshop, the survey design, the protocols, the training courses, the power analyses,
the development of the prevision model and the analyses were all not included in the original budget.
The first year (2010) it was planned to carry out the survey in 8 subareas, all to be densely monitored, but finally,
due to many security and permit problems, the survey included 3 full areas and 3 partial areas. The survey was
carried out by aircrafts not equipped with bubble windows and declinometers.
The second year (2011) it was planned to carry out the survey over 6 areas, all to be densely monitored. An ICCAT
GBYP workshop discussed the details (Anon., 2012), which were adopted by the Steering Committee. Finally,
due to security and permits problems, the survey included only three areas. In this year, following the updated
recommendation of the Steering Committee, the survey was carried out by aircrafts equipped with bubble windows
and declinometers and these tools were used in all following surveys.
The third year (2013) the GBYP Steering Committee requested an extended survey, covering all possible areas in
the Mediterranean Sea. It resulted in 11 different areas, 4 to be densely monitored (these 4 almost overlapping
most of the areas surveyed in previous years) and 7 with less dense transects. At the end, almost all areas were
surveyed, except some parts in three areas, due to security reasons or permit issues. The logistic was extremely
complex, close to impossible.
3174
The fourth year (2015) the GBYP Steering Committee requested again an extended survey, covering all possible
areas in the Mediterranean Sea (about 54.35% of the total surface). It resulted in 11 different areas (partly different
from the previous 11, because of the updated information available on potential bluefin tuna spawning areas), 4 to
be densely monitored (almost overlapping most of the areas surveyed in previous years) and 7 with less dense
transects. The shape of both types of areas was partly different from the ones in 2013, with limited changes for the
areas to be densely monitored. Finally, all areas were surveyed, with the exception of most of the Tunisian airspace,
while security and permits issues affected even this last survey. The logistic was again extremely difficult and
close to impossible.
The companies, pilots and observers were only partly the same during the four survey. This was due to the
administrative structure of ICCAT GBYP (each new Phase is administratively independent from the previous one),
which implies to operate with different Call for tenders and contracts in each Phase; the suspensions in 2012, 2014
and 1016 certainly didn’t help.
Therefore, the GBYP Steering Committee has requested since 2013 a calibration exercise for the spotters, with the
objective to calibrate their sightings and attribute individual CVs for smoothing the additional variance when
elaborating the aerial survey data, but so far it was not possible to carry out any due to serious budget or operational
constraints. In 2015, after many discussions within the Steering Committee and between the Steering Committee
and the GBYP Coordination for trying to find the way for carrying out even a limited calibration exercise, a SWOT
analysis was done (Di Natale, 2016), showing that a calibration is almost impossible for an extended survey like
the ICCAT GBYP one, which includes so many pilots and spotters, while a comprehensive calibration considering
all possible variables is certainly impossible. For this reason, not a single aerial survey for marine animals had so
far a comprehensive one.
As a matter of fact, this is the first time in marine science that an aerial survey is carried out over such a large
proportion of a spawning area which includes so many countries and airspaces, but also a high number of aircrafts,
pilots, professional spotters and scientific spotters, which is certainly an extremely difficult challenge.
At the very end of Phase 5, the Steering Committee requested a new power analyses, which confirmed the results
obtained with the previous two studies, and an extremely detailed analysis of all possible variables included in the
various surveys, including the so-called additional variance
(http://www.iccat.int/GBYP/Documents/ASURVEY/PHASE%205/AERIAL_SURVEY_PHASE5_ANALYSIS
_FINAL_REPORT_1.pdf and
http://www.iccat.int/GBYP/Documents/ASURVEY/PHASE%205/AERIAL_SURVEY_PHASE5_ANALYSIS_
FINAL_REPORT_2.pdf ). Again, the GBYP Steering Committee decided to suspend the survey in 2016.
In 2016 the full GBYP activity has been analysed by two external and independent reviewers (Sissenwine and
Pearce, 2017), including the aerial survey. In the recommendations they stated “A reliable abundance measure for
the Eastern Atlantic Bluefin tuna remains a difficult challenge. It is time to select the method with the best chance
of success and apply enough resources to make it work. Aerial surveys and close kin mark and recapture are the
leading candidates.”
The very short time available between the end of field activities and the SCRS meetings, the necessary time for providing reports and files, the time required for checking all data in details and fixing any possible error or imperfection, made it impossible to present a full report to SCRS in previous years, while full reports were presented in the following years. In 2017 we fully changed the strategy for all reporting issues and for the first time the final report is available for the SCRS Bluefin Tuna Assessment Session. 2. The aerial survey in 2017 In the very first version of the budget proposed by the GBYP Steering Committee for Phase 7, the aerial survey was not included, because it was initially decided to suspend it again, against the opinion of the external reviewers. For the first time, after a specific request by the main funder of the GBYP, the EU (which also considered the strategy approved by the Commission since the beginning of the GBYP, confirmed again by the Commission in 2014), the aerial survey was specifically requested in Phase 7. Therefore, the GBYP Steering Committee reconsidered the activities and the aerial survey was included. The budget originally planned for the aerial survey in 2017 was only 388,000 euro, well below the usual necessary level, due to a general budget restriction for Phase 7. This figure was included in the EU Grant Agreement and was finally approved. This amount included the new survey design, the new protocol, the training course, the survey activities and the analyses of the aerial survey data.
3175
The final cost is to be defined in details, because the administrative part of several contracts is still open on 17
September 2017, and will be better defined after that date. According to the contracts in place, it should be around
406,000 euro, within the flexibility margins of the budget. The survey was carried out having the specific condition
and objective to provide the results to the SCRS Bluefin Tuna Assessment Session, in July 2017.
2.1 Aerial Survey Strategy in 2017
For responding in the necessary manner to the challenge set by the extremely strict schedule for providing the
report, the GBYP Coordination studied and enforced a strategy that was completely different from the one used
so far in all previous survey. At first, the areas for the survey have been set for just the four overlapping areas, so
as to make all data perfectly comparable with those already collected and selected for the same areas. This coverage
was anyway ensuring about 265,625 km2, equivalent to about 10.6% of the full surface of the Mediterranean Sea,
to about 15.15% of both the documented and potential spawning areas and about 46.25% of the documented
Bluefin tuna spawning areas3 (Piccinetti et al., 2013).
Once decided the areas, it was also decided to accelerate as much as possible all the process, setting a precise
schedule for all activities in order to ensure the delivery of the report on time and then going back for each step,
including also some very limited buffer times.
For the very first time, it was decided also to fully change the reporting system, imposing a weekly schedule to
both the companies engaged in the field activities and the team in charge of the analyses, with a further support by
the GBYP coordination team. Even if the schedule has been very tight, it seems that this was a winning strategy.
2.2 Aerial survey design
Even if the GBYP Steering Committee recommended not to change the team in charge of the aerial survey design,
it was considered fair to go for an open Call for tenders, comprehensive of the survey design, the revision of both
the protocol and the sighting forms, the assistance to the training course and the survey data analyses, which was
released on 2 March 2017 (ICCAT GBYP Call for tenders 01/2017, prot. 362/2017). The Call received one bid
that, after the opinion of the Evaluation Committee, was awarded on 15 March 2017 to the same company which
provided the previous survey designs (Alnilam Investigation and Conservation Ltd), including the last one
(Cañadas and Vázquez, 2015a). The contract was issued on 28/03/2017.
The aerial surveys for bluefin tuna in the Mediterranean Sea were all designed using the software DISTANCE
http://www.ruwpa.st-and.ac.uk/distance/ (Buckland et al., 2001), the “industry standard” software for line and
point transect distance sampling based on: the four defined survey areas (survey areas A, C, E and G, see
Figure 1), target survey time available (equivalent to about 32,000 km), time for circling over detected schools to
estimate their size (set at 10%), and time for flying in between lines (set between 10 and 15% depending on the
line separation in each block). Surveys are designed as equal spaced parallel lines. Transect lines were placed
mostly in a north-south direction to be approximately perpendicular to the coast in most blocks (Figure 2);
according to the design, each area has four replicates, while extra additional replicates were included in the design
in case of time or budget availability. The comprehensive ICCAT GBYP aerial survey design for 2017 (Cañadas
and Vázquez, 2017a) is available on line
http://www.iccat.int/GBYP/Documents/ASURVEY/PHASE%207/Aerial_Survey_Design_2017.pdf .
2.3 Aerial survey protocol and sighting forms
The GBYP Protocol for the Aerial Survey on Bluefin Spawning aggregations was issued for the first time in 2015
(http://www.iccat.int/GBYP/Documents/ASURVEY/PHASE%205/AERIAL%20SURVEY%20PROTOCOL%2
0GBYP%202015.pdf ) (Anon., 2015). Due to the various practical issues and situations that were noticed over the
years and particularly those aspects linked to the practical technical part of the survey, it was decided to revise and
update the protocol. The revision work was done in cooperation with the GBYP Coordination Team, checking
every issue and taking into account many improvements; few extreme situations noticed in previous surveys were
deeply discussed and then not included, because of the very high difficulty to face them; it was decided, for all
cases not included in the protocol, to discuss them in real time with the team working on field. The updated protocol
3 The total surface of the Mediterranean Sea is about 2,500,000 km2; the areas where Bluefin tuna spawning is considered extremely unlikely
to occur have been estimated by GBYP in 2014 at about 746,505 km2; therefore, the main documented spawning areas and those where Bluefin
tuna spawning may potentially and occasionally occur were estimated at about 1,753,495 km2, while the most documented Bluefin tuna
spawning areas only are about 574,378 km2 only.
3176
was finally accepted on 16/05/2017 and it is now available on the web
(http://www.iccat.int/GBYP/Documents/ASURVEY/PHASE%207/Aerial_Survey_Protocol_2017.pdf ).
This revision work was also an opportunity for revising and updating the sighting forms used on the aircrafts for
reporting the field activities. The updated form now has 51 field, with a total of 172 entries and it is considered
possibly the most complete available among all aerial surveys for marine animals. The sighting forms is available
on the web (http://www.iccat.int/GBYP/Documents/ASURVEY/PHASE%207/Aerial_Survey_Form_2017.xlsx).
2.4 Aerial survey contracts in 2017
Thanks to the dedicated planning, this year the Call for tender (ICCAT GBYP 02/2017) for carrying out the aerial
survey was issued on 20 March 2017 (prot. 436/2017) and, after the selection process, the contracts were all issued
on 10 May 2017. Three companies were awarded the contracts for various areas: a Spanish company (Grup
AirMed) was awarded for area A (Balearic Sea) and area E (central southern Mediterranean Sea), two Italian
companies (Unimar and Aerial Banners) for areas C and a French company (Action Air Environment/Action
Communication) for area G.
Once awarded the contracts, the ICCAT Secretariat immediately informed all concerned CPCs and assisted all
contractors in all procedures for getting the necessary permits. This work needed a continuous assistance by the
GBYP Coordination, because of the many delicate aspects concerned and many daily difficulties encountered for
various reasons. All companies received the necessary permits, even if some permits had to follow a complex
procedure, due to some peculiar situations4.
2.5 Aerial survey training course
A training course for pilots, professional spotters and scientific observers was organised at the ICCAT Secretariat
in Madrid, on 15 May 2017; it was attended by 22 fellows (for the first time, including the Turkish national
observer), trained by an external expert (Dr. J.A. Vázquez) and by the GBYP Coordinator. During the training
course, the GBYP Coordination carried out an independent assessment of the estimation and identification
capacities of each participant, using a visual tool specifically developed by GBYP. The updated ICCAT GBYP
Protocol for Aerial Survey for Bluefin Tuna Spawning Aggregation (Anon., 2017), the details for filling the
sighting forms and the instructions for the administrative parts were circulated among the contractors immediately
after the course.
Each crew member received a formal ICCAT Identification Card, reporting his/her role in the ICCAT GBYP
Aerial Survey.
2.6 The aerial survey activities in 2017
In 2017, as reported in point 2.1, the reporting strategy was fully changed compared to previous surveys, making
it stricter and much faster. This new approach, which was decided by the GBYP Coordination as the best solution
for facing the extremely strict schedule for providing the aerial survey data for the SCRS Bluefin tuna stock
assessment session in mid-July, was clearly defined in the Call for tenders, linking the deliverables to the weekly
reports and providing the same schedule, shifted by one day, to the team in charge of checking and elaborating the
data. This new approach revealed to be a win-win one, because it allowed to check immediately the reports
provided by each company and team, fix any possible problem in real time and include the verified data into the
system for the final elaboration each week, making the final elaboration and analysis much faster than in the past.
All teams and companies provided the necessary weekly reports on time (in some special cases, the GBYP
Coordination granted one extra day for providing the weekly report).
This year, due to the reduced budget, it was necessary to fit the survey effort with the available budget. Therefore,
the transect length was initially set at 32,000 km, potentially allowing a maximum of four replicates in each area;
according to the previous experience, this could have resulted either in more replicates if the stand-by days are nil
or in less replicates if the stand-by days are higher than the forecast. This design transect length was much higher
than the effective average of the previous surveys (2010, 2011, 2013, 2015) of 21,180 km.
4 ICCAT GBYP acknowledges the very strong cooperation from the Turkish Authorities, which worked hardly for solving various problems
(also linked to the official situation derived to the “emergency status”) and which were able to provide on time the national observer. The
permits for the Turkish area were provided split in several individual permits.
3177
The schedule for beginning the aerial survey was set on 29 May 2017 and the 1st of July was set as the limit for
concluding the field activities. As a matter of fact, the aerial survey field activities initiated on 29 May in area E,
on 30 May in areas A and C, and on June 6 in area G, due to the complexity of the permit procedures and the travel
days. The survey ended on June 14 in area C, on June 26 in areas A and G and on July 2 in area E (both the initial
and ending dates do not include the days needed for reaching the base airport in each area and those for returning
to the home airport). Therefore, the total number of days for effectively carrying out the survey were different in
each area: 29 in area A, 16 in area C, 34 in area E and 21 area G.
The aircrafts were a Partenavia P68V (GBYP ID: ICCAT 1) in area A, a Partenavia P68V (ICCAT ID: ICCAT 2),
and a Partenavia P68C-TC (GBYP ID: ICCAT 6) in area C, a Partenavia P68C (GBYP ID: ICCAT 3)5 in area E
and a CESSNA 337 Sky-master push-pull (GBYP ID: ICCAT 4) in area G. Therefore, we had three areas A, C
and E) covered by high-wings twin side engines aircrafts and one area (G) covered by a high-wings push-pull
engines aircraft. All aircrafts have been equipped with bubble windows, two additional GPS connected to the
computer and declinometers. Each crew had a professional pilot who was also a professional observer, a
professional observer and two scientific observers (except in area G where a scientific observer was substituted by
the Turkish national observer).
The factors affecting the survey in each area were different and Table 1 graphically shows the activities in each
area, including the days on stand-by and the motivations. In total, over 101 days of activities (29 in area A, 16 in
area C, 35 in area E and 21 in area G), the number of days in stand-by was 35, equal to 34.6% against a preliminary
estimation of 25%; including the days with partial activity, then the total reaches 37.5 days, equal to 37.1%. The
percentage of stand-by days by area was 41% in area A, 19% in area C, 41% in area E and 29% in area G. This
high number of days in stand-by was caused by many factors but mostly by the wind (30% in total), that affected
several areas during this period (mostly the Balearic area and the central-southern Mediterranean Sea). This
problem affected also the stabilisation of the thermocline in some parts of these areas, particularly when the wind
continued over several days. Other motivations for the stand-by have been the lack of fuel in area E (a well-known
recurrent problem in Malta over the years which is difficult to solve, due to the lack of Avgas in several airports
close to the area or to the need of a higher rank pilot licence to land in Pantelleria, another airport where Avgas is
available), accounting for 4%, the military activities in area G and some problem to the aircraft in area C (both
accounting for 1% each). As a matter of fact, there was another motivation that only partly appears in Table 1,
and this was the poor visibility in area G, which induced also to adopt a different approach for the strip size; this
limited visibility, generated by a peculiar environmental situation, caused 19% of days of limited operational
activity in the survey in area G.
In previous ICCAT GBYP aerial surveys, the data analyses were available usually at the end of the year (Cañadas
et al., 2010a, 2010b; Cañadas and Vázquez, 2013, 2015b). For the very first time and thanks to the new strategy,
it was possible to get the data elaboration report in real time (Cañadas and Vazquez, 2017b), therefore allowing
for this paper to be presented to the SCRS Bluefin tuna Group two weeks after the conclusion of the field activities.
Figure 3 shows the transects that were effectively surveyed on effort and the tracks off efforts, including the
logistic flights linked to the transects.
In general, in 2017, the aerial survey worked much better than in all previous years, from all points of view and
besides the usual problems. At the beginning it was necessary to discuss and solve the problems with the national
authorities concerned; the problems were related to the permits in three FIRs, the potential security risks in three
areas, the potential problems linked to possible interferences with rescue of migrants in one area and with military
activities and operations in two areas, but at the end everything was solved by the GBYP Coordination working
together side-by-side with the contracted companies concerned. The problems during the field activities were
discussed and solved in real time.
The coverage was very good in all areas (Table 2), for a total of 265,626 km2, even if it was not possible to reach
the total length of the transects set at the beginning, due to several motivations. As a matter of fact, at the end the
final effective transect length was 21,178 km, equal to the average in previous surveys. This evidence confirms
again the right choice of limiting the survey to the four overlapping areas for getting comparable and standardised
results. In 2017, according to the parameters and diagnostics of the detection function, the effective strip width
was defined at 1.4 km in all areas.
5 Due to a problem in the fuel reservoir at the first part of the survey in area C, it was necessary to substitute the aircraft with the reserve one.
3178
The detection functions either using school size as weight or number of animals are identical, and the only thing
changing is the final estimate provided. Therefore, we here refer to it as “the detection function”, even if it was
performed twice.
To explore the effects of both the environmental and the survey factors on the probability of detection function
and therefore on the estimates of abundance, an analysis of each level of the variables was performed and the
results were compared. The final model of the detection function selected was the null model with a Hazard-rate
key function. There was a model with the lowest AIC which had two covariates (subarea and glare30) with a
Hazard-rate key function. However, all diagnostics were better for the null model, the CV of the estimate was
lower and the point estimate was very similar. Therefore, the simplest model was chosen. The Kolmogorov-
Smirnov and the Cramer-von Mises tests performed very well and overall there were no significant differences
between the cumulative distribution function (cdf) and the edf. The q-q plot shows a good agreement between the
cdf and the edf. Table 3 shows the main parameters for the detection function and the results of the diagnostics
tests.
The map showing the distribution of the Bluefin tuna sightings by area, on effort and off effort, during the survey
in 2017, is provided by Figure 4.
The weather and oceanography conditions are extremely important for the aerial survey, particularly in the
Mediterranean Sea, where oceanography factors are essential components for the spawning activities. The general
geography of the Mediterranean area, with so many different coasts and hundreds of isles, naturally creates many
different meteorological situations, over the more that 2.5 million Km2 of the Mediterranean; these conditions may
clearly affect the operational side of the survey. At the same time, the oceanography is quite complex as well, with
effects on the distribution and reproductive biology and behaviour of the Bluefin tuna, and this year it revealed a
further interesting change in the distribution and concentration of the Bluefin tuna schools.
For this reason, the GBYP staff monitored every day SST, waves and wind6, but also salinity, currents and
anomalies, as it was done in the past, quite often checking in real time the maps available on the web by contacting
several people in various sites. Figure 5 shows the situation during the aerial survey in 2017, using three colours:
red for bad weather conditions or for fully unsuitable oceanographic conditions for Bluefin tuna spawning
Piccinetti el al., 2013), yellow for problematic conditions and green for good ones, always taking into account that
these are average estimates for sometimes large areas having mixed situations. Of course, larger the area, greater
the variability within the same area.
Figure 6 shows the average conditions by area. In general, average good conditions were present in 44.6% of the
days during the full extension of the aerial survey period (28 May – 3 July). Negative conditions were there for
23.6% of the time, while problematic conditions affected 31.8% of the days, meaning that during these days there
were zones where it was possible to carry out the survey and zones where it was not possible, sometimes within
the same area. When we take into account the effective days in which the aerial survey was carried out in each
area, then average good conditions were present in 53% of the days during the effective aerial survey period.
Negative conditions were there for 12% of the time, while problematic conditions affected 37% of the days.
Potentially, during the full period the best conditions were in both area C and area G, while taking into account
the effective days, the best conditions were surely in area C, with no negative or problematic days.
Here following there are some observations about the survey activity in each area.
Area A (Balearic Sea)
The aerial survey in area A was initially considered as one of the easiest, due to the fact that the area is well-known
and that the team was almost the same one that was already engaged in all previous surveys. As a matter of fact,
this year the situation was different and much more problematic.
For the first time, we have been notified about some preliminary problems with the permits, even if the aircraft
was Spanish and the airspace is Spanish as well. Possibly, these problems were linked to new security requirements
and controls, but luckily the problems were solved.
6 Oceanographic data were obtained by http://medforecast.bo.ingv.it/ while the daily situation of waves and winds was by
http://isramar.ocean.org.il/isramar2009/wave_model/default.aspx?region=coarse&model=wam
3179
The major problem which limited the survey in the area was the wind, which affected in various ways the survey
in various parts of the area, fully preventing the survey during 12 days (41% of stand-by days against a prevision
of max 25%) (Table 1). Looking in details at the conditions in Area A (Figure 5), it is evident that the waves were
present (at least in some parts of the Balearic area) during many days, while the temperature was also colder than
usual up to June 10, then reaching unusual high temperatures (28°C) on June 22; in both cases, theoretically the
low and high temperatures were not suitable for the Bluefin tuna spawning. According to the data showed on
Figure 6, good conditions were present in Area A just in 29% of the days during the effective aerial survey period;
negative conditions were there for 10% of the time, while problematic conditions affected 69% of the days.
Considering the full extended period (28 May to 3 July), the situation was potentially even worse, with 62% of
problematic conditions, 14% of negative conditions and only 24% days of favourable conditions. Due to the
complex environment, to the above mentioned situations and the presence of the islands (with the implicit complex
oceanography and climate), it is reasonable that the Bluefin tuna spawning season in the Balearic area was much
more concentrated during this part of the season, possibly delayed by a couple of weeks in some areas of the
Balearic Sea, but at the same time quite good, due also to the low presence of jellyfish aggregations.
As concerns the survey activity, besides the unfavourable conditions during several days, it was possible to cover
the full area, even if the adverse conditions and the high number of stand-by days prevented to carry out all replicas
included in the design (Figure 7). Therefore, the total transect length on effort (4,981 km) was lower than the
average in previous surveys but anyway higher than the last survey in 2015 (Table 4). The effective strip width
(1.4 km) was quite reduced, due to the unfavourable conditions and the limited visibility in some days, and this
resulted also in a reduced surface of effective area surveyed (7,017 km2), equal to 11.3% of the total area surface,
the lowest percentage so far. Besides these constraints, the abundance of Bluefin tuna schools in the area was very
high (95), almost the double than the average, and the encounter rate of schools was the highest so far, along with
the density of schools, that are in both cases more than the double of the average. The density of Bluefin tuna was
also very high. Several additional sightings of Bluefin tuna schools were noticed off-effort. The total weight of
Bluefin tuna in area A resulted in 12,693 tons, the highest value so far for this area, about three times than in 2015,
as well as the total abundance of individuals, assessed in 71,520. The CVs of some data were affected by the great
variability of the sightings, sometimes unusually related to single individuals.
The fact that the encounter rates and final estimates are much higher in 2017 than in all previous surveys, even if
with less effort than in most of the previous surveys, indicates that there was a real increase of Bluefin tuna in area
A in 2017 respect the previous 4 surveys.
Area C (southern Tyrrhenian Sea)
The aerial survey in area C was initially considered as one of the easiest, due to the fact that the area is well-known
(possibly the most surveyed in the last three decades) and that the team was almost the same one that has already
been engaged in all previous surveys. As a matter of fact, the forecast was correct. There were no problems with
the flight permit, also because this airspace is fully Italian and the aircrafts were Italian as well.
The only problem that was noticed during the survey happened to the first aircraft and it was a technical problem
in the fuel system, but which caused one day of limited activity, one day in stand-by and the need to substitute the
aircraft with the reserve one, of the same type (it was just a slightly different version). In two other days, the wind
forecast was negative and it was decided not to take-off, finally accounting for a total of 3 stand-by days (18.7%
days of stand-by days against a prevision of max 25%, but the total reach 3.5 days when considering one day of
limited activity, therefore reaching 21.9%) (Table 1). Looking in details at the conditions in Area C (Figure 5), it
seems that even in the two days when the local wind forecast was negative, the sea conditions have been good,
therefore the conditions for the effective period were potentially 100% good (Figure 6). Just after the end of the
aerial survey in area C, we noticed a huge hot water mass in all the southern Tyrrhenian Sea, with very high
temperatures in the SE part, reaching a SST of over 27°C on 17 and 18 June and a peak of 29°C on June 30 and
July 1; these temperatures are sometimes present in the southern Tyrrhenian Sea, but usually in the last part of July
and they are considered too high for allowing the Bluefin tuna spawning. According to the data showed on Figure
6 and considering the full extended period (28 May to 3 July), area C potentially 35% of negative conditions, 11%
of problematic conditions and 54% days of favourable conditions. Therefore, we have been lucky that the survey
in this area has been completed in a very short time, before the negative conditions arrived. Due to the high
temperatures from the second part of June, it is possible that the Bluefin tuna spawning season was between mid-
May to about mid-June. Even in this area, there are no reports of any concentration of jellyfish during the spawning
season.
3180
As concerns the survey activity, due to the good conditions, it was possible to fully complete all transects and
replicas included in the design, and then cover the full area (Figure 8). Therefore, the total transect length on effort
(4,911 km) was higher than in the previous last two surveys, but slightly lower than the average in previous
surveys, due to the higher number of replicas in the first two surveys (Table 5), allowed by a higher budget at that
time. The effective strip width (1.4 km) was quite reduced, due to the parameters used in the detection function
and this resulted in a reduced surface of effective area surveyed (6,918 km2), equal to 12.8% of the total area
surface, the lowest percentage so far. Besides these constraints, the abundance of Bluefin tuna schools in the area
was very high (15), almost the double than the average, and the encounter rate of schools was the second highest
so far, along with the density of schools, that are in both cases about the double of the average. The density of
Bluefin tuna was the highest so far. The total weight of Bluefin tuna in area C resulted in 11,547 tons, the highest
value so far for this area, about three times the average and more than 4 times than the value in 2015, as well as
the total abundance of individuals, assessed in 82,886, which was the double of the average and more than 4 times
than the value in 2015. The spotting of Bluefin tuna schools off-effort was limited to one single school. The CVs
in 2017 have been in general the lowest for the large majority of the data sets obtained by the survey. The increased
presence in area C in 2017 is remarkable, and only comparable with the data obtained in 2013. The high presence
of Bluefin tuna in the southern Tyrrhenian Sea has been also confirmed by the concentration of purse-seines fleets
of several CPCs in this area in the fishing season 2017.
Area E (central-southern Mediterranean Sea)
The aerial survey in area E was always considered a challenge for various reasons: the complex meteorology of
the area, which is subject to rapid changes, the oceanography, which is showing a remarkable variability in the last
decade, the very difficult emergencies of various types in the southern part of the area, the impossibility to access
some airspaces, the potential interferences with military or rescue operations, the low availability of fuel and the
lack of the proper fuel in several airports which can be used for the survey, etc. As a matter of fact, in 2017 we
faced again most of the challenges, but finally it was possible to carry out the survey. At the early beginning, there
was a problem with the flight permit in one airspace, but then it was rapidly solved, thanks to the real-time
intervention of the GBYP Coordination.
The major problem which limited the survey in the area was the wind, which affected in various ways the survey
in various parts of the area, fully preventing the survey during 10 days. Another major problem was the lack of
fuel (Avgas) at the airport in Malta; the late information did not allow to move the aircraft to another airport and
therefore it remained for 4 days in stand-by. In total, even in area E, we had 40% days of stand-by against a
prevision of max 25% (Table 1). Looking in details at the conditions in Area E (Figure 5), it is evident that the
waves were present (at least in some parts of the central-southern Mediterranean Sea) during many days, while the
temperature was also colder than usual up to June 9. It is very possible that the combination of temperatures colder
than usual and winds in some parts of area E prevented the formation of a stable and suitable thermocline, therefore
possibly affecting the suitable spawning conditions for Bluefin tuna in parts of area E. According to the data
showed on Figure 6, good conditions were present in Area E in 50% of the days during the effective aerial survey
period; negative conditions were there for 9% of the time, while problematic conditions affected 41% of the days.
Considering the full extended period (28 May to 3 July) which, in this area, was very close to the effective period,
the situation was potentially even worse, with 43% of problematic conditions, 11% of negative conditions and
46% days of favourable conditions.
As concerns the survey activity, it was possible to cover the full area, but not with the full number of replicas
included in the design (Figure 9), due to the above mentioned constraints, which also forced to use the full
available period of time and even to apply for a short extension. Therefore, the total transect length on effort (6,705
km) was higher that the two last surveys, but slightly lower than the average (Table 6). The effective strip width
(1.4 km) was quite reduced, due to the parameters used in the detection function and this resulted in a reduced
surface of effective area surveyed (9,466 km2) was the lowest of all surveys, equal to 10.1% of the total area
surface, the lowest percentage so far. The abundance of Bluefin tuna schools in the area was very low (9), even if
it was three times the one reported in 2015, but it was anyway about 42% of the average; is mirroring this, as well
as the density of schools. The density of Bluefin tuna was the second lowest so far, even if about four times higher
than in 2015. The estimated total weight of Bluefin tuna in area E resulted in 4,457 tons, higher than in the last
two surveys, but about 42% of the average. The total abundance of individuals, assessed in 36,927, was about just
27% of the average, even if it was more than three times higher than the last survey in 2015. The CVs in 2017
have been variable, depending on the type of parameter taken into account, but they have been all much lower than
in 2015. The observations of Bluefin tuna in off-efforts conditions in area E have been minimal. The low
availability of Bluefin tuna in several parts of area E in 2017 was also confirmed by the movements of various
purse-seiners fleets, which had left the area after the first days of unsuccessful searching.
3181
Area G (Levantine Sea)
The aerial survey in area G, again, was always considered a challenge for various reasons: the oceanography,
which is usually characterised by an early increase of the SST and even in the water masses, which usually induce
an early spawning activity for the Bluefin tuna in this area, the weather conditions, because the wind can affect
some areas in several days, the complex procedures to get some flight authorisations, the impossibility to access
some airspaces, the uncertain definition of some areas where bilateral agreements are lacking, the potential
interferences with military operations, etc. As a matter of fact, in 2017 we faced again all the challenges, but finally
it was possible to carry out the survey, with the exception of two small areas for security motivations (this exclusion
was discussed before initiating the survey and it was decided that the survey would not be too much affected). At
the early beginning, there was a problem with the flight permits, but then it was rapidly solved, thanks to the real-
time intervention of the GBYP Coordination and the strong cooperation of the Turkish authorities and the EU
authorities. Furthermore, for the first time, and thanks again to the support of the Turkish authorities, it was
possible to have all spotters on board duly trained by the GBYP, including the national observer. Even is some
operational problems happened during the survey, the survey in this area in 2017 should be regarded as a positive
one.
The major problems which limited the survey in the area were the meteorological conditions, because not only
there were 5 days of wind which prevented the flights, but there was also a peculiar situation, which affected many
days of survey: the very limited visibility at sea due to the brumes. This anomalous situation was induced by the
high temperatures, the lack of wind and the very high humidity, all combined together; even if the visibility
communicated by the control towers in the various airports was more than 10 km, at sea the real visibility was
usually close to 1.2 km. After discussing in real time with the team in charge of the analyses, it was decided to go
on, accurately noting the real visibility on the sighting forms. Anyway, this specific situation caused four days of
limited survey activity. The military activities in the area resulted in one day of stand-by. In total, in area G, we
had 28.6% of stand-by days against a prevision of max 25% (Table 1), but without considering the other 19% of
days with reduced activity. Looking in details at the conditions in Area G (Figure 5), it is evident that the waves
were present (at least in some parts of the Levantine Sea) during just 4 days, while the temperature was also slightly
colder than usual at the beginning of the season, but then it went over >27°C on June 16 for several days for then
reaching higher temperatures at the very end of June and in the first days of July. According to the data showed
on Figure 6, good conditions were present in Area G in 57% of the days during the effective aerial survey period;
negative conditions were there for 29% of the time, while problematic conditions affected 14% of the days.
Considering the full extended period (28 May to 3 July), the situation was potentially slightly worse, with 54% of
favourable conditions, 11% of problematic conditions and 35% of negative conditions.
As concerns the survey activity, it was possible to cover almost the full area included in the design (Figure 10),
except the two small areas excluded by the permit and one transect for logistic problems. Therefore, the total
transect length on effort (4,581 km) was the highest so far, about the double of the average (Table 7). The effective
strip width (1.4 km) was quite reduced, due very low visibility in several days (as described above) and to the
parameters used in the detection function, and this resulted in a reduced surface of effective area surveyed (6,453
km2), which was anyway the second highest and in the average. The abundance of Bluefin tuna schools in the area
was the highest so far (191), about the double of the average and more than 8 times higher than the results in 2015,
as mirrored also by the density of schools. The density of Bluefin tuna was also the highest so far, eleven times
higher than in 2015 and more than the double of the average (but data are not available for the first two years).
The estimated total weight of Bluefin tuna in area G resulted in 3,157 tons, much higher than in the last two
surveys, and close to the average. The total abundance of individuals, assessed in 159,939, was again the highest
so far and more than the double of the average. The CVs in 2017 are quite low compared with the previous ones,
except for some in 2010. The observations of Bluefin tuna in off-efforts conditions in area G have been many,
further confirming the abundance of the species in this area in 2017.
3. Discussion
As reported in the previous paragraphs, the relevant improvements in the survey and data reporting and analysis
strategy in 2017 resulted in much more coherent and better estimates. As said before, the exclusive use of
overlapping areas for carrying out the survey, the mandatory weekly reporting, the weekly check of the data, the
improved training course, the improved and updated protocol resulted all together in the easier elaboration of the
data and in a very short time for getting the final results.
3182
The combined data for the four areas surveyed in 2017 are shown on Table 8 and it is very clear that the aerial
survey in 2017 worked very well, even taking into account the reduced budget availability, which imposed a
reduced number of replicas compared to years when the budget was much higher, and considering also the
unfavourable weather conditions in some areas, which limited both the operations and the effective strip width.
Besides the practical problems, most of which are unpredictable but always within the usual alea of a wide field
activity, the activity this year has been a win-win one.
The results show that the total survey area was equal to 265,627 km2, for a final effective transect length of 21,178
km and a total effective area searched of 29,834 km2. This last number is just the result of the reduced effective
strip width (1.4 km, imposed mostly by the reduced visibility in one area), because, as a matter of fact, the searched
area was much larger. The number of Bluefin tuna schools detected on effort (91) has been the highest so far,
confirming a good presence of the species.
The abundance of schools (387) was one of the highest so far, almost the same than the highest value (388)
registered in 2011 and much higher than the average. The encounter rate of schools (0.0043) was the highest so
far, about the double than the average. The density of schools (1.457/1000 km2) has been the second highest so
far, well over the average. The mean weight of the schools was 82.3 tons, below the average, for a high presence
of young spawners. The density of animals (1.304 km2) has been the second highest, even in this case over the
average. The main parameters, the total weight (31,855 tons) and the total abundance of fish (n=346,272) have
been both the second highest so far, well over the average (Figure 11 graphically shows the results in all surveys
for the density of fish and schools, the total abundance and the total weight).
As described in the previous chapter, a shifting of presence was noticed in 2017, about the opposite of what it was
noticed in 2006 (Di Natale, 2008) and, at that time, the anomaly was motivated with a drastic change in the Eastern
Mediterranean Transient (EMS). As a matter of fact, the presence in the central-southern Mediterranean Sea
decreased in a remarkable manner, while a parallel increase was noticed in the southern Tyrrhenian Sea (but also
in the other areas) (Table 9). This fact was noticed not only by the ICCAT GBYP aerial survey, but it was
confirmed also by the adaptation of the fishing strategy of the purse-seiner fleets in the same days. What are the
conditions which caused this shifting of presence? We cannot provide a response at the moment, even if many
oceanographers have been contacted, because the analyses of the data collected in 2017 will be available at the
end of the year. At the same time, as mentioned in chapter 2.6, some important anomalies have been noticed in
real time in 2017. The SST showed anomalous high data in several parts of the central and western Mediterranean
Sea, while the Levantine Sea and the Gulf of Sirte did not show hot temperatures (Figures 12 and 13). This is a
fact, but if this was one of the causes which induced a different concentration of Bluefin tunas during the spawning
season in 2017 is not defined.
Was this shifting in presence someway linked to the about two-weeks anticipated arrival of Bluefin tunas in all
traps (both in the eastern Atlantic Ocean or in the Mediterranean Sea) in 2017? Even in this case, besides the fact
that the western and central areas of the Mediterranean had some high temperature anomalies in mid-April, we
don’t have any answer at the moment.
In general, the ICCAT GBYP aerial survey is an extremely challenging activity, not only for the many factors
which can bias the observations in different ways (Di Natale, 2016), but also for the many difficulties for operating
in a so large area and so many aerial spaces managed by so many countries. This activity implies also that several
aircraft and spotters have to be used at the same time, because of the short time frame in a normal spawning season
for the Bluefin tuna. All these factors combined together were never tested in a same survey.
In addition to these factors, there are others linked to the survey methodology itself (DISTANCE), which is
currently considered the best available for marine species. Line transect sampling assumes that detection on the
transect line itself is certain, while, in the reality, the random factor plays a very important role, which is not fitted
in the function. In this last survey, the situation was better than most of the previous ones and therefore the fitted
detection function, combining all areas, fitted much better than in the past (Figure 14 shows the fitted detection
function in 2017 and Figure 15 shows the Q-Q plot; Figure 16 shows the fitted detection function in 2010, 2011,
2013 and 2015).
For the aerial surveys, in general and in the reality, it is not possible to assume that the detection function is certain,
because the flight speed means that some schools available to be “sampled” will inevitably not be detected (so-
called perception bias). In addition, Bluefin tuna spends a variable part of its time beneath the surface and, in this
case, it is unavailable for the detection (the so-called availability bias) (Quilez Badía et al., 2016). This latter
specific factor will be possibly corrected by using the data obtained by several electronic tags that were deployed
3183
by ICCAT GBYP and other entities, at least for having the percentage of time at surface of the fish that went into
the Mediterranean spawning areas when the aerial survey was in place. Table 10 and Figure 17 show the average
percentage of time spent by the Bluefin tuna between 0 and 10 meters7 depth during the various months. the data
are obtained from the PSATs deployed by ICCAT GBYP between 2011 and 2015 (Tensek et al., 2017), updated
with the data obtained from the electronic tags deployed in 2016 and up to 2017; the data taken into account are
only those related to the Mediterranean Sea. Applying these percentages (accounting for the average May-June-
July, equal to 60.03, which correspond to an increasing factor of about 160%) to the two main parameters, the total
weight and the total number of fish, then the estimated values should be 50,968 tons and 554,035 fish in just the
four areas taken into account. The trend is the same but the levels are different. Estimates of abundance from these
surveys are therefore implicitly underestimates (minimum estimates) even though a detection function has been
fitted to correct for fish missed within the survey strip.
The appropriateness of these estimates as indices of abundance for the future depends on a number of factors
including: timing of surveys (should be usual the same), areas surveyed (now we are working only on standardised
overlapping areas), and stability of availability and perception biases. Availability and perception bias can be
reasonably assumed as almost stable over time, but the knowledge of the distribution in time and space of the
Bluefin tuna throughout the Mediterranean Sea is still incomplete and subject to variables which are difficult to be
detected. To minimise natural variation in using survey estimates as indices of abundance over time and therefore
detecting trends, surveys in future years should ideally be conducted always in the same areas and at the same time
of year.
Again, it is very clear that, due to the high interannual variability between the four main spawning areas in the
Mediterranean Sea, any type of fishery-independent trend cannot be based on one single area, because the bias
could be huge and fully undefined, while it should be based on at least the four main areas.
4. Acknowledgements
This work has been possible thanks to the field work conducted by the following crew members of the various
aircrafts:
- Grup AirMed (Areas A and E): María García de Prado Olaizola and Jorge San Antolín San Frutos (P),
Luis Navarro Martínez and Carlos Dos Santos Silva (PS), José Antonio Vázquez Bonales and Mónica
Pérez Gil, Helder Fernandes Araujo and José Martínez Cedeira (SS).
- Unimar and Aerial Banners (Area C): Francesco Orrico, Daniele Mercurio and Francesco Ruggiero (P);
Salvatore De Martino Mario Piscino and Vincenzo Severino (PS); Adriano Mariani, Simone Serra,
Andrea Fusari (SS).
- Action Air Environnement/Action Communication: Patrick Féron (P) ; Khalifa Zariohi (PS) ; Cihan
Toslak (NO/SS) ; Léa Lecomte (SS).
This work was carried out under the provision of the ICCAT Atlantic-Wide Research Programme for Bluefin Tuna
(GBYP), funded by the European Union, by several ICCAT CPCs, the ICCAT Secretariat and by other entities.
The contents of this paper do not necessarily reflect the point of view of ICCAT or of the other funders, which
have not responsibility about them, neither do them necessarily reflect the views of the funders and in no ways
anticipate the Commission’s future policy in this area.
7 Even in this case, the limits are very prudential, because with a good water transparency, the spotters are quite used to estimate the schools
of Bluefin tunas at greater depths.
3184
Bibliography
Anonimous, 2012, 2011 GBYP Workshops on Aerial Surveys, and Operational Meetings on Biological Sampling
and on Tagging of Bluefin Tuna (Madrid, Spain, February 14-18, 2011). Coll. Vol. Sci. Pap., ICCAT, 68
(1): 1-65.
Anonimous, 2015, ICCAT GBYP Aerial Survey Protocol 2015. http://www.iccat.int/GBYP/en/asurvey.htm .
Anonimous, 2017, ICCAT Atlantic-wide Programme for Bluefin tuna (GBYP). Aerial Survey Protocol 2017: 1-
17. http://www.iccat.int/GBYP/Documents/ASURVEY/PHASE%207/Aerial_Survey_Protocol_2017.pdf
Arena P., 1980, Observation aeriennes sur la distribution et le comportement du Thon rouge, Thunnus thynnus
(L.), de la Mer Tyrrhenienne. XXVII Congr. Assem. Plen. CIESM, Cagliari.
Arena P., 1981, Osservazioni sulle concentrazioni e sulla pesca del Tonno e dell’Alalunga nelle zone di mare
meridionali. Quad.Lab.Tecn.Pesca., 3(1), suppl.: 77-79.
Arena P., 1982a, Biologia, ecologia e pesca del tonno (Thunnus thynnus L) osservati in un quinquennio nel Tirreno
meridionale. Atti Conv. UU.OO: sottop. Ris.Biol.Inq.Marino, Roma: 381-405.
Arena P., 1982b, Composizione demografica dei branchi di tonno (Thunnus thynnus, L.) durante il periodo
genetico, con indicazioni utili alla individuazione dello stock di riproduttori che affluiscono nel Mar
Tirreno. Atti Conv. UU.OO. sottop. Ris. Biol. Inq. Marino, Roma.
Arena P., 1982c, La pêche à la senne tournante du thon rouge, Thunnus thynnus (L.), dans les bassins maritimes
occidentaux italiens. Collect. Vol. Sci. Pap. ICCAT, 17(2): 281-292.
Arena P., 1985, La pesca del tonno in Sicilia. Atti Conv.Pesca e Trasf. Prod. Itt. Siciliani, Trapani: 23-28.
Arena P., 1986a, Sullo stato e le caratteristiche della pesca in Italia dei grandi Teleostei pelagici (Tonno, Alalunga
e Pescespada). Rapp. Min.Mar.Merc., miméo: 1-17.
Arena P., 1986b. Pesca dei grandi Scombroidei e degli Xifioidei nei mari Italiani. Nova Thalassia, 8 (3): 657-658.
Arena P., 1988a, Risultati delle rilevazioni sulle affluenze del tonno nel Tirreno e sull’andamento della pesca da
parte delle “tonnare volanti” nel triennio 1984-1986. MMM-CNR, Atti Seminari UU.OO. Resp.Prog. Ric.,
Roma: 273-297.
Arena P., 1988b, Rilevazioni e studi sulle affluenze del tonno nel Tirreno e sull’andamento della pesca da parte
delle “tonnare volanti” nel quadriennio 1984-1988. Report to: ESPI, Ente Siciliano per la Promozione
Industriale, Palermo, 1-55, I-XI.
Arena P., 1990, Rilevazioni e studi sulle caratteristiche e lo stato delle risorse di Tonno e sugli andamenti della
pesca (Relazione sulla prosecuzione 1987-89). Report to: ESPI, Ente Siciliano per la Promozione
Industriale, Palermo, 1-58.
Arena P., Cefali A., Potoschi A., 1979, Risultati di studi sulla biologia, la distribuzione e la pesca dei grandi
scombro idei nel Tirreno meridionale e nello Ionio. X(4): 329-345.
Arena P., Cefali A., 2002, Composizione demografica dei branchi di tonno, Thunnus thynnus (L.) durante il
periodo genetico, con indicazioni utili alla individuazione dello stock di riproduttori che affluiscono nel
Mar Tirreno. Atti Conv. UU: OO. Ris. Biol. Inq. Marino, Roma: 425-442.
Basson M., Farley J.H., 2014, A standardised abundance index from commercial spotting data of Southern Bluefin
Tuna (Thunnus maccoyii): random effects to the rescue. PLOS One,
https://doi.org/10.1371/journal.pone.0116245
Bauer R., Bonhommeau S., Brisset B., Fromentin J.-M., 2015a, Aerial surveys to monitor bluefin tuna abundance
and track efficiency of management measures. Marine Ecology Progress Series, 534: 221-234 . Publisher's
official version: http://doi.org/10.3354/meps11392, Open Access version:
http://archimer.ifremer.fr/doc/00281/39192/
Bauer R.B., Fromentin J.-M., Demarcq H., Brisset B., Bonhommeau S., 2015b, Co-occurrence and habitat use of
Fin Whales, Striped Dolphins and Atlantic Bluefin tuna in the Northwestern Mediterranean Sea. PLOS
ONE, DOI: 10-1371/journal.pone.0139218.
Bonhommeau S., Farrugio H., Poisson F., Fromentin J.M., 2010, Aerial surveys of bluefin tuna in the Western
Mediterranean sea: retrospective, prospective, perspective. ICCAT Coll. Vol. Sci. Pap., 65 (3): 801-811.
Bower R., Fromentin J.M., Bonhommeau S., Demarcq H., 2014, Estimating Juvenile Bluefin Tuna Abundance
through Aerial Surveys in the Northwestern Mediterranean. Quebec 2014, 144 Annual Meeting American
Fisheries Society. https://afs.confex.com/afs/2014/webprogram/Paper14274.html
Buckland S.T., Anderson D.R., Burnham K.P., Laake J.L., Borchers D.L., Thomas L., 2001. Introduction to
distance sampling: estimating abundance of biological populations. Oxford University Press, Oxford.
3185
Cañadas A., Hammond P., Vázquez J.A., 2010a, GBYP Data Recovery Plan. Elaboration of 2010 data from the
aerial survey on Bluefin tuna spawning aggregations. Final report, 27/09/2010.
http://www.iccat.int/GBYP/Documents/ASURVEY/PHASE%201/Aerial%20survey%20data%20analysis
%20final%20report%202010.pdf
Cañadas A., Hammond P., Vázquez J.A., 2010b, GBYP Data Recovery Plan. Elaboration of 2010 data from SST
and the aerial survey on Bluefin tuna spawning aggregations. Final report, 03/12/2010.
http://www.iccat.int/GBYP/Documents/ASURVEY/PHASE%201/Aerial%20survey%20spatial%20mode
lling%20analysis%20final%20report%202010.pdf
Cañadas A., Vázquez J.A., 2013, Short-term contract for assessing the feasibility of a large-scale aerial survey on
Bluefin tuna spawning aggregations in all the Mediterranean Sea for obtaining useful data for operating
modelling purposes. Final Report, 13 January 2013. ICCAT GBYP, Phase 3.
http://www.iccat.int/GBYP/Documents/ASURVEY/PHASE%203/Aerial_Survey_Feasibility_Study_Pha
se3.pdf .
Cañadas A., Vázquez J.A., 2015a, Short term contract for the aerial survey design of the Atlantic-wide research
programme for Bluefin tuna (ICCAT GBYP Phase 5 - 2015). http://www.iccat.int/GBYP/en/asurvey.htm .
Cañadas A., Vázquez J.A., 2015b, Elaboration of 2015 data from the aerial survey on Bluefin spawning
aggregations (ICCAT GBYP Phase 5 - 2015). Preliminary Report, 14/09/2015: 1-70.
Cañadas A., Vázquez J.A., 2017a, Aerial survey design of the Atlantic-wide Research Programme on Bluefin tuna
(ICCAT GBYP Phase 7 – 2017). Report, 17/03/2017: 1-68.
http://www.iccat.int/GBYP/Documents/ASURVEY/PHASE%207/Aerial_Survey_Design_2017.pdf .
Cañadas A., Vázquez J.A., 2017b, Atlantic-wide Research Programme on Bluefin Tuna (ICCAT GBYP- Phase 7
– 2017). Elaboration of data from the aerial survey on spawning aggregations. Report, 18/07/2017: 1-25.
Cowling, A. O'Reilly, J., 1999, Background to review of aerial survey: 3rd draft. 1999 Aerial Survey Workshop,
Caloundra: 1-61.
Cowling A., Millar C., Polacheck T., 1996, Data analysis of the aerial surveys (1991–1997) for juvenile southern
bluefin tuna in the Great Australian Bight. Rep. RMWS/96/4, 87 p. Recruitment Monitoring Program,
CSIRO Division of Marine Research, GPO Box 1538, Hobart 7001, Australia
Cram D.L., Hampton I., 1976, A proposed aerial/acoustic strategy for pelagic fish stock assessment. J. Cons. int.
expl. Mer., 37: 91–97.
Di Natale A., 2008, I cambi climatici nel Mediterraneo: modificazioni della struttura della pesca e possibili
adattamenti. Cambiamenti Climatici: Conferenza Nazionale, MATTM, 20/07/2007, Brindisi: 1-22.
Di Natale A., 2011, ICCAT GBYP. Atlantic-wide Bluefin Tuna Research Programme 2010. GBYP Coordinator
Detailed Activity Report for 2009-2010. Coll. Vol. Sci. Pap., ICCAT, 66(2): 995-1009.
Di Natale A., 2016, Tentative SWOT analysis for the calibration of ICCAT GBYP aerial survey for Bluefin tuna
spawning aggregations. SCRS/2015/143. Collect. Vol. Sci. Pap., ICCAT, 72 (6): 1463-1476.
Di Natale A., Idrissi M., 2012, 011 ICCAT-GBYP Atlantic-wide Research Programme for Bluefin Tuna (GBYP)
Coordination. Detailed Activity Report for Phase 2. Coll. Vol. Sci. Pap., ICCAT, 68 (1): 176-207.
Di Natale A., Idrissi M., 2013a, ICCAT-GBYP Atlantic-wide Research Programme for Bluefin Tuna 2012. GBYP
Coordination detailed activity report on Phase 2 (last part) and Phase 3 (first part). Coll. Vol. Sci. Pap.,
ICCAT, 69 (2): 760-802.
Di Natale A., Idrissi M., 2013b, ICCAT GBYP aerial survey: juveniles versus spawners, a SWOT analysis of both
perspectives. Coll. Vol. Sci. Pap., ICCAT, 69 (2): 803-815.
Di Natale A., Tensek S., 2016, ICCAT Atlantic-wide Research Programme for Bluefin Tuna (GBYP): activity
report for the last part of Phase 4 and the first part of Phase 5 (2014-2015). Collect. Vol. Sci. Pap., ICCAT,
72 (6): 1477-1530.
Di Natale A., Idrissi M., Justel Rubío A., 2014a, ICCAT-GBYP activities for improving knowledge of bluefin
tuna biological and behavioral aspects (ICCAT-GBYP phases 1 to 3). Coll. Vol. Sci. Pap., ICCAT, 70 (1):
249-270.
Di Natale A., Idrissi M., Justel Rubío A., 2014b, ICCAT Atlantic-wide Research Programme for Bluefin Tuna
(GBYP) activity report for 2013 (extension of Phase 3 and first part of Phase 4). Coll. Vol. Sci. Pap.,
ICCAT, 70 (2): 459-498.
Di Natale A., Cañadas A., Vázquez Bonales J.A., Tensek S. and Pagá García A., 2016, ICCAT GBYP aerial survey
for bluefin tuna spawning aggregations in 2015. Preliminary report. Collect. Vol. Sci. Pap., ICCAT, 72 (6):
1553-1577.
Everson E.J., Bravington M.V., Farley J.H., 2011, A mixed effects model for estimating juvenile southern bluefin
tuna abundance from aerial survey data. ICCAT GBYP Worshop on Aerial Survey, Madrid, 1-27.
3186
Farley J., Bennet B., 2008, A bird’s eye view of Southern Bluefin Tuna. CSIRO, Wealth from Oceans: 4 p.
Fortuna C., Holcer D., Filidei E. Jr, Donovan G., Tunesi L., 2011, First cetacean aerial survey in the Adriatic Sea:
summer 2010. In: Seventh Meet ACCOBAMS Sci Committee. ACCOBAMS-SC7/2011/Doc06, 29–31
March 2011, Monaco.
Fortuna M.C., Kell T.L., Holcer D., Canese S., Filidei E.Jr., Mackelworth P., Donovan G,, 2014, Summer
distribution and abundance of the giant devil ray (Mobula mobular) in the Adriatic Sea: baseline data for
an integrated management framework. Scientia Marina, 78 (2): doi:
http://dx.doi.org/10.3989/scimar.03920.30D
Fromentin J.-M. 2001, Interim Report of STROMBOLI - EU-DG XIV project 99/022. 60 pp.
Fromentin J.-M., Farrugio H., Deflorio M., De Metrio G. 2003, Preliminary results of aerials surveys of bluefin
tuna in the Western Mediterranean Sea. Collect. Vol. Sci. Pap. ICCAT, 55(3): 1019-1027.
Fromentin J.M., Bonhommeau S., Brisset B., 2013, Update of the index of abundance of juvenile bluefin tuna in
the western Mediterranean Sea until 2011. Collect. Vol. Sci. Pap., ICCAT, 69 (1): 454-641.
Grierson J., 1949, Air whaler.Sampson Low, Morston & Co. Ed., London: 1-243.
Hammond P.S., Berggren P., Benke H., Borchers D.L., Collet A., Heide-Jørgensen M.P., Heimlich S., Hiby A.R.,
Leopold M.F., Øien N., 2002, Abundance of harbour porpoise and other cetaceans in the North Sea and
adjacent waters. Journal of Applied Ecology 39, 361–376
Hiby L., Lovell P., 1998, Using aircraft in tandem formation to estimate abundance of harbor porpoise. Biometrics,
54, 1280–1289.
Kessel S.T., Gruber S.H., K. S. Gledhill K.S., Bond M.E., Perkins R. G., 2013, Aerial Survey as a tool to estimate
abundance and describe distribution of a Carcharhinid species, the Lemon Shark, Negaprion brevirostris.
Journal of marine Biology,
Lauriano G., Panigada S., Casale P., Pierantonio N., Donovan G.P., 2011, Aerial survey abundance estimates of
the loggerhead sea turtle Caretta caretta in the Pelagos Sanctuary, northwestern Mediterranean Sea. Mar.
Ecol. Prog. Ser., 437: 291−302.
Lauriano G., Pierantonio N., Kell L., Cañadas A., Donovan G., Panigada S., 2017, Fishery-independent surface
abundance and density estimates of swordfish (Xiphias gladius) from aerial surveys in the Central
Mediterranean Sea. Deep-Sea Research, Part II: Topical studies in Oceanography, 141: 102-114.
Lutcavage M., Krauss S., Hoggar W., 1997, Aerial survey of giant bluefin tuna, Thunnus thynnus, in the Great
Bahama bank, Straits of Florida, 1995. Fishery Bulletin, 95: 300-310.
Lutcavage M., Newlands N., 1999, A strategic framework for fishery-independent aerial assessment of bluefin
tuna. Collect. Vol. Sci. Pap. ICCAT, 49: 400-402.
Marsh H., Sinclair D.F., 1989, Correcting for visibility bias in strip transect aerial surveys of aquatic fauna, Journal
of Wildlife Management, vol. 53 (4): 1017–1024.
Newlands N.T., Lutcavage M.E., Pitcher T.J., 2007, Atlantic bluefin tuna in the Gulf of Maine, II: precision of
sampling designs in estimating seasonal abundance accounting for tuna behaviour. Envir. Biol. Fish.,
Palka D., 2011, U.S. Survey aerial survey experience. ICCAT GBYP Workshop on Aerial Survey, Madrid, 10-11
February 2011: 1-32.
Panigada S., Lauriano G., Burt L., Pierantonio N., Donovan G., 2011, Monitoring Winter and Summer Abundance
of Cetaceans in the Pelagos Sanctuary (Northwestern Mediterranean Sea) Through Aerial Surveys. PLOS
ONE, http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0022878
Panigada S., Lauriano S., Donovan G., Pierantonio N., Cañadas A., Vazquez J.A., Burt L., 2017, Estimating
cetacean density and abundance in the Central and Western Mediterranean Sea through aerial surveys:
Implications for management. Deep-Sea Research, Part II: Topical studies in Oceanography, 141: 102-114.
Piccinetti C., Di Natale A., Arena P., 2013, Eastern bluefin tuna (Thunnus thynnus, L.) reproduction and
reproductive areas and seasons. Collect. Vol. Sci. Pap. ICCAT, 69(2): 891-912.
Polacheck T., Pikitch E., Lo. N. 1996, Evaluation and recommendations for the use of aerial surveys in the
assessment of Atlantic bluefin tuna. Coll. Vol. Sci. Pap. ICCAT, 68(1):61–68.
Quílez Badía G., Tensek S., Di Natale A., Pagá García A., Cañadas A., Kitakado T., Kell L.T., in press, A
consideration of additional variance in the Mediterranean aerial survey based on electronic satellite tags.
SCRS/2015/146, Collect. Vol. Sci. Pap., ICCAT, 72 (6): 1544-1552.
Rivas L.R., 1978. Aerial surveys leading to 1974-1976 estimates of the numbers of spawning giant bluefin tuna
(Thunnus thynnus) migrating past the western Bahamas. ColI. Vol. Sci., ICCAT (7): 301-312.
3187
Rouyer T., Bonhommeau S., Brisset B., Fromentin J.-M., 2016, Aerial surveys of bluefin tuna in the Western
Mediterranean Sea: an operational fishery-independent abundance index for juvenile? SCRS/2016/153
(withdrawn).
Rouyer T., Brisset B., Bonhommeau S., Fromentin J.-M., in press, Update of the abundance index for juvenile fish
derived from aerial survey of bluefin tuna in the western Mediterranean Sea. SCRS/2017/044, Collective
Volume of Scientific Papers.
Sissenwine M., Pearce J., 2017, Second review of the ICCAT Atlantic-wide Research Programme for Bluefin tuna
(ICCAT GBYP Phase 6-2016). Collect. Vol. Sci. Pap., ICCAT, 73 (7): 2340-2423.
Tensek S., Di Natale A., Pagá García A., 2017, ICCAT GBYP PSAT tagging: the first five years. Collect. Vol.
Sci. Pap., ICCAT, 73 (6): 2058-2073.
3188
Table 1. General overview of the field activities by area, including the motivations for stand-by.
Table 2. Areas, number and total length of transects and number of sightings of Bluefin tuna for each surveyed
sub-area.
Area
Area
(km2)
Number
of
transects
Length of
transects
on effort
(km)
Number of
observations
(after
truncation)
Detection
Function
Number of
observations
(after
truncation)
Abundance
estimate
A 61,933 26 4,981 40 22
C 53,868 25 4,911 16 15
E 93,614 30 6,705 10 9
G 56,211 55 4,581 61 45
Total 265,626 136 21,178 127 91
Table 3. Parameters and diagnostics of the detection function for the ICCAT GBYP Aerial Survey in 2017.
Average
probability
of detection
(p)
Effective
strip width
(esw)
(km)
Chi-
square
test
K-S
test
(p)
Cramer-von Mises
test (unweighted)
(p)
0.1803 0.704 0.7252 0.8689 0.8721
2017
AREA 28 29 30 1 2 3 4 5 6 8 10 11 13 14 15 16 18 19 20 21 22 23 24 25 26 27 28 29 30 1 2 3
A (Balearic Sea)
C (S.Tyrrhenian Sea)
E (C.S.Mediterranean)
G (Levantine Sea)
stand-by motivation wind visibility lack of fuel aircraft problems military activities
others travel limited activity
MAY
31
JULY
7 9 12 17
JUNE
3189
Table 4. Survey details for the surveys carried out so far in Area A (Balearic Sea). All data are only related to the
same overlapping surface and to on-effort results, excluding the off-effort data.
Year 2010 2011 2013 2015 2017 Total
(sum)
Total
(mean)
Survey area (km2) 61,933 61,933 61,933 61,933 61,933 309,665 61,933
Transect length (km) 6,118 7,838 6,807 4,109 4,981 29,852 5,970
Effective strip width x2 (km) 2.96 1.36 3.00 3.9 1.4 2.52
Area searched (km2) 18,130 10,660 20,398 15,961 7,017 72,166 14,433
% coverage 29.3 17.2 32.9 25.8 11.3 23.3
Number of schools ON effort 8 10 10 6 22 56 11.2
Abundance of schools 25 58 30 23 95 231 46
%CV abundance of schools 55.4 35.9 36.1 43.4 30.8
Encounter rate of schools 0.0013 0.0013 0.0015 0.0014 0.0044 0.00198
%CV encounter rate 54.5 33.8 35.1 41.1 25.9
Density of schools (1000 km-2) 0.402 0.938 0.490 0.372 1.531 0.747
%CV density of schools 55.4 35.9 36.1 43.4 30.8
Mean weight (t) 131.25 122.43 194.1 160.7 133.9 148.462
%CV weight 6.2 19.2 23.8 11.7 34.9
Mean cluster size (animals) 678.1 611 825 754 717
%CV abundance 27.9 26.0 11.0 33.6
Density of animals (km-2) 0.636 0.299 0.307 1.155 0.599
%CV density of animals 45.4 44.5 44.7 39.7
Total weight (t) 3,587 4,371 3,539 4,712 12,693 5,780
%CV total weight 56.5 46.2 40.6 42.0 40.9
L 95% CI total weight 1,251 1,807 1,624 2,132 5,848
U 95% CI total weight 10,285 10,577 7,710 10,414 27,551
Total abundance (animals) 39,399 18,542 19,002 71,520 37,116
%CV total abundance 45.4 44.5 44.7 39.7
L 95% CI total abundance 16,540 7,913 8,195 33,620
U 95% CI total abundance 93,850 43,445 44,060 152,141
3190
Table 5. Survey details for the surveys carried out so far in Area C (southern Tyrrhenian Sea). All data are only
related to the same overlapping surface and to on-effort results, excluding the off-effort data.
Year 2010 2011 2013 2015 2017 Total
(sum)
Total
(mean)
Survey area (km2) 53,868 53,868 53,868 53,868 53,868 269,340 53,868
Transect length (km) 8,487 8,826 2,791 2,739 4,911 27,754 5,550
Effective strip width x2 (km) 2.96 1.36 3.00 3.9 1.4 2.52
Area searched (km2) 25,150 12,004 8,364 10,640 6,918 63,076 12,615
% coverage 46.7 22.3 15.5 19.8 12.8 23.4
Number of schools ON effort 6 10 10 3 15 44 8.8
Abundance of schools 12 45 64 13 57 38
%CV abundance of schools 45.7 33.4 34.3 62.0 28.8
Encounter rate of schools 0.0007 0.0011 0.0036 0.0009 0.0031 0.0016
%CV encounter rate 44.6 31.2 33.1 60.5 23.6
Density of schools (1000 km-2) 0.217 0.833 1.196 0.239 1.058 0.709
%CV density of schools 45.7 33.4 34.3 62.0 28.8
Mean weight (t) 124.17 38.87 173.5 190.0 202.5 145.808
%CV weight 5.6 44.4 22.1 19.9 21.9
Mean cluster size (animals) 733 291 1,285 1,533 1,453 1,059
%CV abundance 36.5 30.7 17.0 19.0 17.2
Density of animals (km-2) 0.182 0.242 1.536 0.366 1.539 0.773
%CV density of animals 59.2 45.3 38.3 64.9 33.3
Total weight (t) 1,596 1,917 11,370 2,665 11,547 4,387
%CV total weight 46.9 54.9 40.8 65.1 35.5
L 95% CI total weight 652 661 5,161 802 5,829
U 95% CI total weight 3,904 5,557 25,049 8,856 22,874
Total abundance (animals) 9,797 13,059 82,763 19,708 82,886 41,643
%CV total abundance 59.2 45.3 38.3 64.9 33.3
L 95% CI total abundance 3,187 5,446 39,399 5,958 43,597
U 95% CI total abundance 30,016 31,317 173,860 65,192 157,580
3191
Table 6. Survey details for the surveys carried out so far in Area E (central-southern Mediterranean Sea). All data
are only related to the same overlapping surface and to on-effort results, excluding the off-effort data.
Year 2010 2011 2013 2015 2017 Total
(sum)
Total
(mean)
Survey area (km2) 93,614 93,614 93,614 93,614 93,614 468,070 93,614
Transect length (km) 13,137 10,192 4,381 2,566 6,705 36,981 7,396
Effective strip width x2 (km) 2.96 1.36 3.00 3.9 1.4 2.52
Area searched (km2) 38,930 13,862 13,129 9,969 9,446 85,335 17,067
% coverage 41.6 14.8 14.0 10.6 10.1 18.2
Number of schools ON effort 29 45 20 3 9 106 21.2
Abundance of schools 63 304 135 20 44 113
%CV abundance of schools 31.5 24.1 34.8 58.0 36.4
Encounter rate of schools 0.0022 0.0044 0.0046 0.0008 0.0013 0.0029
%CV encounter rate 29.9 21.0 33.6 56.3 32.4
Density of schools (1000 km-2) 0.678 3.246 1.447 0.213 0.466 1.210
%CV density of schools 31.5 24.1 34.8 58.0 36.4
Mean weight (t) 110.14 118.05 11.0 50.2 102.3 78.338
%CV weight 33.9 19.2 66.0 99.5 51.2
Mean cluster size (animals) 1,015 1,715 361 507 848 889
%CV abundance 19.0 21.5 67.3 97.9 33.2
Density of animals (km-2) 0.787 5.566 0.522 0.108 0.395 1.476
%CV density of animals 37.8 32.3 75.7 113.8 49.9
Total weight (t) 7,681 37,851 1,517 1,093 4,457 10,520
%CV total weight 47.1 32.2 74.6 115.2 63.4
L 95% CI total weight 3,155 20,342 390 75 1,413
U 95% CI total weight 18,698 70,432 5,899 15,857 14,062
Total abundance (animals) 73,676 521,085 48,884 10,126 36,927 138,140
%CV total abundance 37.8 32.3 75.7 113.8 49.9
L 95% CI total abundance 35,741 279,620 12,363 727 14,559
U 95% CI total abundance 151,880 971,060 193,280 141,020 93,662
3192
Table 7. Survey details for the surveys carried out so far in Area G (Levantine Sea). All data are only related to
the same overlapping surface and to on-effort results, excluding the off-effort data.
Year 2010 2011 2013 2015 2017 Total
(sum)
Total
(mean)
Survey area (km2) 56,211 56,211 56,211 56,211 224,844 56,211
Transect length (km) 3,790 2,081 859 4,581 11,311 2,827
Effective strip width x2 (km) 2.96 3.00 3.9 1.4 2.81
Area searched (km2) 11,231 6,236 3,335 6,453 27,256 6,814
% coverage 20.0 11.1 5.9 11.5 12.1
Number of schools ON effort 33 12 2 45 92 23
Abundance of schools 150 108 22 191 118
%CV abundance of schools 28.1 39.7 70.9 23.5
Encounter rate of schools 0.0087 0.0058 0.0015 0.0098 0.0081
%CV encounter rate 26.3 38.7 69.5 16.6
Density of schools (1000 km-2) 2.674 1.924 0.399 3.398 2.099
%CV density of schools 28.1 39.7 70.9 23.5
Mean weight (t) 63.621 4.0 9.0 16.5 23.280
%CV weight 12.7 40.2 66.7 31.5
Mean cluster size (animals) 336 600 809 582
%CV abundance 36.7 66.7 31.9
Density of animals (km-2) 0.646 0.239 2.756 1.214
%CV density of animals 54.1 97.3 40.1
Total weight (t) 10,507 440 220 3,157 3,581
%CV total weight 32.1 56.5 97.3 39.3
L 95% CI total weight 5,643 151 25 1,495
U 95% CI total weight 19,561 1,285 1,965 6,669
Total abundance (animals) 36,316 13,448 154,939 68,234
%CV total abundance 54.1 97.3 40.1
L 95% CI total abundance 12,995 1,506 72,366
U 95% CI total abundance 101,490 120,070 331,731
3193
Table 8. Results for all ICCAT GBYP aerial surveys in all overlapping areas combined.
Year 2010 2011 2013 2015 2017 Total
(sum)
Total
(mean)
Survey area (km2) 265,627 209,416 265,627 265,627 265,627 1,288,135 265,627
Transect length (km) 31,532 26,856 16,060 10,272 21,178 105,898 21,173
Effective strip width x2 (km) 2.96 1.36 3.00 3.9 1.4 2.52
Area searched (km2) 93,442 36,525 48,127 39,904 29,834 166,041 33,208
% coverage 35.2 17.4 18.1 15.0 11.2 12.89
Number of schools ON effort 76 65 52 14 91 298 59.6
Abundance of schools 250 388 338 78 387 288
%CV abundance of schools 22.8 19.9 21.5 38.9 20.2
Encounter rate of schools 0.0024 0.0024 0.0032 0.0014 0.0043 0.0028
%CV encounter rate 20.2 11.6
Density of schools (1000 km-2) 0.942 1.852 1.274 0.295 1.457 1.086
%CV density of schools 22.8 19.9 21.5 38.9 23.4
Mean weight (t) 87.9 101.1 22.6 272.2 82.3 113.212
%CV weight 16.8 27.5 51.0 41.4 19.2
Mean cluster size (animals) 791 1,275 582 1,548 895 1018
%CV abundance 18.6 37.3 18.5 40.5 17.0
Density of animals (km-2)
2.7388 0.702 0.234 1.304 1.245
%CV density of animals 29.9 29.4 39.1 25.9
Total weight (t) 23,371 44,139 16,866 8,690 31,855 24,984
%CV total weight 25.6 28.7 30.3 35.3 26.7
L 95% CI total weight 14,243 25,315 9,343 4,398 19,018
U 95% CI total weight 38,347 76,964 30,447 17,169 53,355
Total abundance (animals)
573,543 186,505 62,284 346,272 292,151
%CV total abundance 29.9 29.4 39.1 25.9
L 95% CI total abundance 321,620 105,320 28,766 209,816
U 95% CI total abundance 1,022,800 330,270 134,860 571,473
3194
Table 9. Results for all ICCAT GBYP aerial surveys in all overlapping areas and in total in 2017.
Year A C E G Total
(sum)
Total
(mean)
Survey area (km2) 61,933 53,868 93,614 56,211 265,627
Transect length (km) 4,981 4,911 6,705 4,581 21,178
Effective strip width x2 (km) 1.4 1.4 1.4 1.4 1.4
Area searched (km2) 7,017 6,918 9,446 6,453 29,834
% coverage 11.3 12.8 10.1 11.5 11.2
Number of schools ON effort 22 15 9 45 91 22.8
Abundance of schools 95 57 44 191 387 96.8
%CV abundance of schools 30.8 28.8 36.4 23.5 20.2
Encounter rate of schools 0.0044 0.0031 0.0013 0.0098 0.0043
%CV encounter rate 25.9 23.6 32.4 16.6 11.6
Density of schools (1000 km-2) 1.531 1.058 0.466 3.398 1.457
%CV density of schools 30.8 28.8 36.4 23.5 23.4
Mean weight (t) 133.9 202.5 102.3 16.5 82.3
%CV weight 34.9 21.9 51.2 31.5 19.2
Mean cluster size (animals) 754 1,453 848 809 895
%CV abundance 33.6 17.2 33.2 31.9 17.0
Density of animals (km-2) 1.155 1.539 0.395 2.756 1.304
%CV density of animals 39.7 33.3 49.9 40.1 25.9
Total weight (t) 12,693 11,547 4,457 3,157 31,855
%CV total weight 40.9 35.5 63.4 39.3 26.7
L 95% CI total weight 5,848 5,829 1,413 1,495 19,018
U 95% CI total weight 27,551 22,874 14,062 6,669 53,355
Total abundance (animals) 71,520 82,886 36,927 154,939 346,272
%CV total abundance 39.7 33.3 49.9 40.1 25.9
L 95% CI total abundance 33,620 43,597 14,559 72,366 209,816
U 95% CI total abundance 152,141 157,580 93,662 331,731 571,473
Table 10. Mean percentage of time spent in the Mediterranean Sea between the sea surface 10 meters depth, of all
Bluefin tuna that were tagged by GBYP between 2011 and 2016 (data up to 2017), by month (from: Tensek et al.,
2017, updated).
Month Mean time spent at 0-10 m depth (%)
January 42,33%
February 38,59%
March 41,78%
April 50,12%
May 63,03%
June 56,02%
July 61,05%
August 54,03%
September 49,30%
October 47,41%
November 43,21%
December 40,72%
Total 53,02%
3195
Figure 1. The four areas identified for the aerial survey in 2017. They correspond to the overlapping areas in all
previous surveys and to the most important Bluefin tuna spawning areas in the Mediterranean Sea.
Figure 2. The transect design for the four areas to be surveyed by GBYP in 2017. Each area has four replicates,
while extra replicates are not showed on this figure.
3196
Figure 3. ICCAT GBYP transects flown on effort and off efforts, including logistics flights, in 2017.
Figure 4. Distribution of the sightings of Bluefin tuna on and off effort during the ICCAT GBYP Aerial Survey
for spawning aggregations in 2017.
3197
Figure 5. Graphic representation of potential SST and waves high conditions during the aerial survey activity in
2017, in correlation with suitable situations for Bluefin tuna spawning and the aerial survey. Combined data are
shown by the “index”.
Figure 6. Graphic representation of potential positive, problematic or positive meteo-marine conditions in the
various areas and corresponding percentages, for both the full period of the survey and the effective survey period
in each area.
2017
AREA 28 29 30 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 1 2 3
SST
A (Balearic Sea)
C (S.Tyrrhenian Sea)
E (C.S.Mediterranean)
G (Levantine Sea)
Waves
A (Balearic Sea)
C (S.Tyrrhenian Sea)
E (C.S.Mediterranean)
G (Levantine Sea)
Index
A (Balearic Sea)
C (S.Tyrrhenian Sea)
E (C.S.Mediterranean)
G (Levantine Sea)
SST (°C) waves (m) index [AVG(temp & waves)]
1 <20 >2 1-1,5 (if any temp or waves =1)
2 20-22 1-2 2-2.5
3 >22-27 <1 3
1 >27
JULYJUNEMAY
full period
days in red days in yellow days in green
A 5 14% A 23 62% A 9 24%
C 13 35% C 4 11% C 20 54%
E 4 11% E 16 43% E 17 46%
G 13 35% G 4 11% G 20 54%
effective period
days in red days in yellow days in green
A 3 10% A 20 69% A 8 28%
C 0 0% C 0 0% C 16 100%
E 3 9% E 14 41% E 17 50%
G 6 29% G 3 14% G 12 57%
3198
Figure 7. Transects realized in Area A (Balearic Sea), and sightings of Bluefin tuna on and off effort.
Figure 8. Transects realized in Area C (southern Tyrrhenian Sea), and sightings of Bluefin tuna on and off effort.
3199
Figure 9. Transects realized in Area E (central-southern Mediterranean Sea), and sightings of Bluefin tuna on and
off effort.
Figure 10. Transects realized in Area G (Levantine Sea), and sightings of Bluefin tuna on and off effort.
3200
Figure 11. Graphic plots of the main results of the five ICCAT GBYP Aerial surveys on Bluefin tuna spawning
aggregations for the density of schools and animals (top), the total abundance of fish (middle) and the total weight
in tons (bottom).
3201
Figure 12. Few examples about the evolution of the SST in the Mediterranean Sea between the last part of May
to the last days of June 2017. It is very clear that the SST reached high levels in the western Mediterranean, in the
Ligurian and Tyrrhenian Sea, in the Adriac Sea and in the western Strait of Sicily, much higher than usual compare.
At the same time, the southern Mediterranean Sea and the Levantine Sea were not hot as usual in this season.
3202
Figure 13. Few examples about the evolution of the SST anomalies in the Mediterranean Sea between the last part
of May to the last days of June 2017. It is very clear that the SST was much higher tha usual in the western
Mediterranean, in the Ligurian and Tyrrhenian Sea, in the Adriac Sea and in the western Strait of Sicily compared
to the average. At the same time, the southern Mediterranean Sea and the Levantine Sea were not so hot.
3203
Figure 14. Detection function, scaled to 1.0 at zero perpendicular distance, and histograms of observed sightings
during the ICCAT GBYP Aerial Survey in 2017.
Figure 15. Q-Q plot for the ICCAT GBYP Aerial Survey in 2017.
Distance
De
tectio
n p
rob
ab
ility
0 2000 4000 6000 8000
0.0
0.2
0.4
0.6
0.8
1.0
0.0 0.2 0.4 0.6 0.8 1.0
0.0
0.2
0.4
0.6
0.8
1.0
Empirical cdf
Fitte
d c
df
3204
Figure 16. Detection function for 2010, 2011, 2013 and 2015, scaled to 1.0 at zero perpendicular distance, and
histograms of observed sightings.
Figure 17. Mean percentage of time Bluefin tuna spent on the surface (0-10 m depth) by months, according to the
data provided by electronic pop-up tags deployed by ICCAT GBYP. The data are related only to the Mediterranean
Sea, for the tags deployed between 2011 and 2016, with data up to 2017. The orange line represents the mean
percentage of time spent on the surface (0-10 m depth) throughout the year (from: Tensek et al., 2017, updated).
0%
10%
20%
30%
40%
50%
60%
70%%